Statistical power in two-level models: A tutorial based on Monte Carlo simulation.

Matthias G Arend, Thomas Schäfer

February 2019 Psychol Methods

Synopsis of Social media discussions

This collection of discussions reflects widespread appreciation for the article's practical guidance, with examples like mentions of simulation tools such as R and Jamovi, and references to key papers like Arend & Schäfer (2019). The tone varies from informative to enthusiastic, emphasizing both the tutorial's utility and its potential to improve research quality.

A
Agreement
 Moderate agreement

Most individuals acknowledge the importance of the tutorial and its contributions to understanding power analysis in multilevel models, supported by references to key papers and methods.

I
Interest
 High level of interest

Discussions show a high level of excitement and curiosity, with references to practical tools like R and Jamovi, indicating strong engagement with the topic.

E
Engagement
 Moderate level of engagement

People reference specific concepts such as Monte Carlo simulations, effect sizes, and sample size guidelines, demonstrating active involvement and deeper understanding.

I
Impact
 Moderate level of impact

The discussions highlight that the article influences best practices in statistical analysis and encourages adopting simulation tools for research design.

Social Mentions

YouTube

2 Videos

Twitter

21 Posts

Metrics

Video Views

4,364

Total Likes

161

Extended Reach

18,688

Social Features

23

Timeline: Posts about article

Top Social Media Posts

Posts referencing the article

Understanding Power Analysis in Hierarchical Models with Monte Carlo Methods

Understanding Power Analysis in Hierarchical Models with Monte Carlo Methods

Estimating statistical power in two-level models is complex due to hierarchical data structures. This tutorial demonstrates Monte Carlo simulations and the SIMR method for effective power analysis, providing practical guidelines for sample sizes and effect detection.

ADRIPS

November 13, 2023

3,389 views

Understanding Two-Level Regression with R and SPSS

Understanding Two-Level Regression with R and SPSS

This video covers conducting two-level regression using R, Stata, Mplus, or SPSS. It introduces a three-step procedure for analysis, including centering predictors and establishing final models. Explore the practical applications as highlighted in the referenced article for effective multilevel modeling.

ADRIPS

January 15, 2024

975 views

  • Marie Delacre
    @mdelacre1 (Twitter)

    RT @ADRIPS_comm:
    view full post

    January 10, 2024

    9

  • Camille Grasso
    @grasso_camille (Twitter)

    RT @ADRIPS_comm:
    view full post

    December 19, 2023

    9

  • Laurence Chaby
    @chab_laurence (Twitter)

    Qu’est-ce qu’une analyse de puissance ? Calculer sa taille d’échantillon-cible avec Jamovi, G*Power et R, par @brice_beffara https://t.co/HaBCN0fV4i #Statistics #jamovi
    view full post

    November 21, 2023

    1

  • Psychologie robuste et fiable
    @sci_psy (Twitter)

    RT @ADRIPS_comm:
    view full post

    November 15, 2023

    9

  • Brice Beffara
    @brice_beffara (Twitter)

    RT @ADRIPS_comm:
    view full post

    November 15, 2023

    9

  • le mec chiant
    @idaho20 (Twitter)

    RT @ADRIPS_comm:
    view full post

    November 13, 2023

    9

  • LAPPS - équipe ENOSIS
    @LAPPS_UP8 (Twitter)

    RT @ADRIPS_comm:
    view full post

    November 13, 2023

    9

  • Gaëlle Marinthe
    @GaelleMarinthe (Twitter)

    RT @ADRIPS_comm:
    view full post

    November 13, 2023

    9

  • Fanny Ollivier
    @fanny_oll (Twitter)

    RT @ADRIPS_comm:
    view full post

    November 13, 2023

    9

  • JB Légal
    @jblegal (Twitter)

    RT @ADRIPS_comm:
    view full post

    November 13, 2023

    9

  • ADRIPS
    @ADRIPS_comm (Twitter)


    view full post

    November 13, 2023

    18

    9

  • Nicolas Sommet
    @nicolas_sommet (Twitter)

    @CForestier_PhD A great piece is: Arend & Schäfer (2019,
    view full post

    August 9, 2022

    3

  • Niclas Kuper
    @niclas_kuper (Twitter)

    @georg_henning @AStenling @GinetteLafit Generally important to note that power for even moderate cross-level interaction effects is often low with typical sample sizes. See also this excellent paper: https://t.co/7Xen7dDxbQ
    view full post

    February 14, 2022

    4

  • Ryan
    @Ryanzoriaa (Twitter)

    RT @ddlcoppersmith: @dp_moriarity @eisenlohr_moul Citations: “Statistical power in two-level models: A tutorial based on Monte Carlo simula…
    view full post

    October 19, 2020

    1

  • Daniel Coppersmith
    @ddlcoppersmith (Twitter)

    @dp_moriarity @eisenlohr_moul Citations: “Statistical power in two-level models: A tutorial based on Monte Carlo simulation”: https://t.co/VUqz5veTRP “Number of Subjects and Time Points Needed for Multilevel Time-Series Analysis”: https://t.co/XDdLTCHlig
    view full post

    October 19, 2020

    7

    1

  • Maurizio Sicorello @mauriziosicorello@fediscience
    @MLSicorello (Twitter)

    @eisenlohr_moul @AleksaKaurin Favorite paper on LMM power analysis: https://t.co/H5JuqBw39e On standardized effect sizes in LMMs: https://t.co/2hipFErWGT. But although non-LMM, this is the reason I prefer unstandardized/POMP scores LMM: https://t.co/ib7FPcmZQp
    view full post

    October 19, 2020

    2

  • Leslie Brick
    @LeslieBrickPhD (Twitter)

    @JTWaddell7 Arend and Schafer (2019) is a handy paper that includes some power tables based on MC simulaton for two-level models with varying N and T. Also has R code to run your own simulations. https://t.co/6k9pkBoVUM
    view full post

    September 21, 2020

    5

  • Daniel Gucciardi
    @DanielGucciardi (Twitter)

    @pdakean "Statistical power in two-level models: A tutorial based on Monte Carlo simulation" inc demonstration with R package SIMR & guidelines for sufficient sample sizes (80% power) for various effect sizes & sizes of the variance components => https://t.co/l6dBBmeyLW
    view full post

    January 19, 2019

    3

  • Dr. Danielle Molnar
    @DPHWB_Lab (Twitter)

    RT @DanielGucciardi: Nice paper "Statistical power in two-level models: A tutorial based on Monte Carlo simulation" inc demonstration with…
    view full post

    January 17, 2019

    2

  • Mari Todd
    @Marirunriderow (Twitter)

    RT @DanielGucciardi: Nice paper "Statistical power in two-level models: A tutorial based on Monte Carlo simulation" inc demonstration with…
    view full post

    January 16, 2019

    2

  • Daniel Gucciardi
    @DanielGucciardi (Twitter)

    Nice paper "Statistical power in two-level models: A tutorial based on Monte Carlo simulation" inc demonstration with R package SIMR & guidelines for sufficient sample sizes (80% power) for various effect sizes & sizes of the variance components => https://t.co/IIFLws3RXR
    view full post

    January 16, 2019

    5

    2

Abstract Synopsis

  • This text explains that estimating statistical power in two-level models, which analyze hierarchically structured data, is complex due to variance at two levels and predictors at both levels.
  • It introduces a hands-on tutorial using Monte Carlo simulations and the SIMR method to perform pre- and post-hoc power analyses, including guidance on setting standardized input parameters and interpreting results.
  • The study provides practical rules of thumb for sample sizes and detectable effect sizes, indicating that moderate effects can be identified with certain sample sizes, while small effects, especially at the higher level, remain difficult to detect with typical cluster numbers.